Nonpharmacological Therapy for Atrial Fibrillation: An Historical Overview
David D. Spragg
Hugh Calkins
I have noticed that after the heart proper, and even the right auricle were ceasing to beat and appeared on the point of death, an obscure movement, ondulation-palpation had clearly continued in the right auricular blood itself for as long as the blood was perceptibly imbued with warmth and spirit.
—William Harvey, Exercitatio Anatomica De Motu Cordis et Sanguinis in Animalibus, 1628
Atrial fibrillation (AF) is the most common sustained tachyarrhythmia encountered by physicians. The prevalence of AF in patients over the age of 65 is approximately 6%, and approaches 10% in patients over the age of 85 (1). As the median age of the population in the United States becomes older, the epidemiologic burden of AF in this country will likely increase. Currently approximately 2.2 million people in the United States have AF (1). AF, although typically not a life-threatening arrhythmia per se, is associated with increased risk of stroke (2), heart failure, and increased mortality. The stroke risk in patients with AF, for instance, is increased between five- and sevenfold compared to similar patients without AF (3,4).
Therapy for AF can be divided into two major paradigms—rate control and rhythm control. Rate control, as the name implies, focuses exclusively on preventing an uncontrolled, rapid ventricular response rate in the setting of AF. Strategies to achieve rate control typically include either pharmacological agents to slow conduction through the atrioventricular (AV) node (i.e., beta-blockers or calcium-channel blockers), or ablation of the AV junction and implantation of a permanent pacemaker. Large prospective randomized trials have validated rate control as a reasonable option
in patients with AF, particularly in terms of overall mortality (5,6). However, such a strategy does nothing to reduce the stroke risk and loss of AV synchrony seen in patients with AF, and as such, represents a suboptimal strategy in many patients.
in patients with AF, particularly in terms of overall mortality (5,6). However, such a strategy does nothing to reduce the stroke risk and loss of AV synchrony seen in patients with AF, and as such, represents a suboptimal strategy in many patients.
The second paradigm, rhythm control, has historically involved the use of antiarrhythmic medications and/or DC cardioversion from AF into sinus rhythm. Antiarrhythmic medications used for the maintenance of sinus rhythm include class I and class III agents. Randomized prospective data have demonstrated that amiodarone, compared to other class III and to class I medications, is the most effective antiarrhythmic drug to prevent AF (7,8). Long-term therapy with amiodarone is imperfect, however, due both to limited efficacy and to attendant end-organ toxicities. Recurrence rates in patients treated with amiodarone are approximately 35% (7). As important, amiodarone has dose-dependent effects on thyroid, liver, and pulmonary function. In patients treated with DC cardioversion alone (i.e., without the suppressive effects of antiarrhythmic medications) AF recurrence is predictably high, with nearly 66% of cardioverted patients developing recurrent AF within 15 months (9). In part because of the limitations of effective and safe pharmacological therapy for AF suppression, clinicians have sought nonpharmacological interventions to achieve rhythm control. Two major modalities—surgical intervention and catheter-based intervention for the treatment of AF—have developed over the past 25 years. The current chapter reviews these two modalities, with a particular emphasis on the historical evolution behind each, and on the seminal discoveries made during that evolution that inform our current practices in 2007.
Surgical Treatment for Atrial FibrillationIn 1959, Moe hypothesized that AF was due to multiple randomly propagating reentrant waves in the atrium, suggesting that functional reentry was the mechanism underlying fibrillation (10, 11, 12). In subsequent work by Allessie et al. (13, 14, 15), Moe’s hypothesis was confirmed. AF was demonstrated to require at least six to eight circulating reentrant wavefronts. Maintenance of AF depended both on a critical atrial mass and on conduction velocity and refractory periods in the atrial tissue to support functional reentry.
Twenty years after Moe’s hypothesis, James Cox et al. began a series of revolutionary clinical experiments to cure AF surgically (16, 17, 18, 19). Initial surgical interventions focused on electrical isolation of the left atrium (20). Although this procedure resulted in AV synchrony and controlled ventricular response rates, patients still were at risk of atrial thrombosis and embolism, due to ongoing fibrillation in the isolated atrium. In the mid-1980s, Cox et al. developed a surgical procedure designed to eliminate AF rather than isolate it—the MAZE procedure. It was based on a series of incisions in the left and right atria that would facilitate sinus node to AV junction conduction while at the same time preventing propagation of fibrillatory wave fronts. The initial MAZE-I procedure resulted in unacceptable high rates of chronotropic incompetence, and was modified to the MAZE-II procedure. Because the MAZE-II procedure was technically difficult to perform, even for its inventors, it was further modified into the MAZE-III procedure (21). The MAZE-III involved a complex set of incisions in both the left and right atria. These incisions included circumferential lesions around the left- and right-sided pulmonary veins (PVs), lesions from the left superior PV (LSPV) to the mitral valve (MV) annulus, left atrial appendage removal, and a series of lesions in the right atrium (19,21).
The MAZE-III procedure has become the standard by which other surgical interventions for AF have been judged. Short-, intermediate-, and long-term data are available from Cox et al. for patients undergoing the MAZE-III (22, 23, 24), both for lone AF and in conjunction with other cardiovascular procedures (24). In the 5 years between 1987 and 1992, 47 patients underwent either the MAZE-I or MAZE-II procedure. From then until 2000, Cox et al. have performed 308 MAZE-III procedures. Operative mortality for this group was 2.9%, with short-term AF or atrial flutter (AFL) recurrence rates of 37% (20). These high recurrence rates, however, were not indicative of long-term procedure failure. Rather, they likely represent atrial irritation and changes in atrial electrophysiological properties unique to the postoperative period. In a review of 8.5 years of follow-up, Cox et al. reported a 98% success rate, defined as freedom from any AF (23). Half of the treatment failures by surgery alone were successfully treated with antiarrhythmic drugs, for an overall cure rate of 99%. The stroke rate in this population, anticoagulated for the initial postoperative period but then off of systemic anticoagulant therapy, was 0.7%. A recent report on the 15-year follow-up of these patients shows that the results continue to be durable, with a cure rate of >95% (24).
Despite the high success rates and strikingly low stroke rates in patients undergoing the MAZE-III, widespread adoption of the procedure has been limited. In part this has been due to concerns about perioperative sequelae including the need for permanent pacemaker implantation and the loss of atrial mechanical function. Cox et al. have reported that up to 15% of patients undergoing the MAZE-III have subsequently had pacemakers implanted postoperatively. Most of these patients, however, had antecedent sinus node dysfunction that became clinically apparent after cessation of AF, rather than acute injury to the AV conduction system during surgery. Atrial function appears to be preserved in the Cox MAZE-III cohort, with >93% preservation of biatrial transport function. The technically challenging nature of the surgery, rather than lack of efficacy or attendant complications, appears to be the principal factor limiting widespread adoption of the full MAZE-III procedure. In 1996, Cox et al. developed a minimally invasive version of the MAZE (25). This procedure, in which cryoablation is used to form the atrial lesion set through a limited thoracotomy, has yielded promising short-term results, but does not yet have the long-term follow-up of the MAZE-III procedure.
Other surgical interventions for AF have been developed (20,26,27). Some are of purely historical interest. In 1985, Guiraudon described a unique surgical treatment of atrial fibrillation, in which a corridor of intact atrial tissue connecting the sinus node and the AV junction was isolated (20). Although this procedure allowed for ventricular rate control by the sinus node, it necessarily left the bulk of atrial tissue in fibrillation. As such, the corridor procedure did nothing to reduce thromboembolic risk or atrial mechanical dysfunction due to AF, and was ultimately abandoned.
Non-MAZE surgical approaches are also currently in practice. Gillinov et al. at the Cleveland Clinic have presented results from a series of 513 patients undergoing surgical ablation of AF using an epicardial radiofrequency (RF) ablation system (Atricure Inc, West Chester, OH) (28). In this series, patients with preoperative AF presenting for cardiac surgery underwent concomitant PV isolation and LA appendage removal. Most patients were referred for valvular surgery. Although the perioperative AF incidence was 57%, 84% of patients were free of AF after 1 year. No strokes were observed during long-term follow-up, and 4% of patients required pacemaker implantation.
The number of randomized clinical trials investigating the efficacy of MAZE-III or maze-related procedures is limited. Reston et al. identified four such trials, in which MV surgery in conjunction with AF ablative lesion application was compared to MV surgery alone (29). They found that restoration of sinus rhythm (81% vs. 17%) and prevention of subsequent stroke (0% vs. 5.8%) were strongly favored by MAZE-III or maze-related surgery. These procedures had no significant impact on overall mortality. Pacemaker implantation was higher in patients undergoing AF surgical intervention than in patients undergoing MV surgery alone (3.9% v. 1.5%).
In summary, there has been a 25-year history of surgical intervention for AF. The gold standard for this procedure has been the Cox MAZE-III procedure. This intervention has demonstrated the virtual elimination of both AF recurrence and long-term stroke incidence. The morbidities associated with cardiac surgery, however, and the complexity of the lesion set has led both surgeons and interventional cardiologists to investigate other nonpharmacological interventions that approach the treatment results of the MAZE-III. Catheter-based attempts to recreate the MAZE-III lesion set have evolved over the last 15 years, and form the basis of current endovascular procedures to treat AF. The evolution of these procedures is discussed below.
Catheter-Based Treatment for Atrial FibrillationCatheter ablation for atrial tachyarrhythmias is a relatively recent phenomenon (30, 31, 32, 33, 34, 35). The propagation of electrical activation from atria to ventricles over myocardial fibers was originally described in 1883 (36). Nearly a century later, Scheinman et al. described the first catheter-based ablation procedure—His bundle interruption for the control of ventricular response rates to refractory supraventricular tachycardias (37). Over the last 25 years, catheter ablation techniques have become standard, curative therapy for AVNRT (32), accessory pathway ablation (33,35), and ablation of macroreentrant atrial flutter (34). Although ablation of the AV junction has long been accepted as a palliative treatment for AF, curative catheter-based therapy has evolved rapidly since the early 1990s. Initial work focused on linear and MAZE-like lesions sets in the right (38,39), right and left (40), and left atria (41). More recently, the importance of AF triggers (particularly those located in the PVs) has been recognized and targeted (42).
Linear Ablation Strategies for Treating Atrial Fibrillation
Initial ablation attempts to cure atrial fibrillation focused on linear lesions confined to the right atrium. Between 1994 and 1996, Haissaguerre et al. investigated the effects of linear lesion sets in patients with symptomatic, drug-refractory AF (40). Forty-five initial patients were studied and followed over the long term. Patients initially underwent right atrial ablation only, with either a single ablation line from SVC to IVC over the atrial septum, or multiple lines (longitudinal and transverse) to compartmentalize the right atrium. The procedure led to stable sinus rhythm in 18 of 45 patients (40%) during the procedure. Sustained AF was inducible in 40 of 45 patients, however, and 19 patients underwent repeat ablation of left- or right-sided atrial flutter or focal atrial tachycardia. After a follow-up period of 11 ± 4 months, only six patients were free of AF off antiarrhythmic drugs, with another nine patients free of AF on a previously ineffective medication (overall success of 33%). Nine of 45 patients had significant improvement of their symptom burden with the aid of an antiarrhythmic medication, whereas the remaining 21 of 45 patients had no appreciable effect from RA-only ablation. After 26 ± 5 months of follow-up (39), there was a further reduction in therapeutic benefit,
with seven previous responders (either cure or significant reduction in AF burden) reverting to frequent AF. Successful results with RA-only lesions were seen in only 17 of 45 patients.
with seven previous responders (either cure or significant reduction in AF burden) reverting to frequent AF. Successful results with RA-only lesions were seen in only 17 of 45 patients.
Other investigators have prospectively attempted curative lesion sets confined to the right atrium. Natale et al. studied 18 patients with symptomatic, drug-refractory AF (43). Although the lesion sets varied somewhat among the patients (seven with two intercaval lesions, ablation of the cavotricuspid isthmus, and an anterior RA line; 11 with a single intercaval line, a septal line, and cavotricuspid isthmus ablation), the results were generally poor. After a follow-up period of 22 ± 11 months, only five of 18 patients remained free from atrial arrhythmia recurrence. Most of the 13 recurrences occurred within 2 months of the procedure. The particular lesion set did not predict procedure efficacy. Thus, although linear ablation confined to the right atrium to cure AF is attractive from a technical and safety standpoint, multiple trials with intermediate and long-term follow-up have shown it to be a largely ineffective procedure (39,44).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
